The present investigation reveals that severe inflammatory nociceptive insults occurring in neonatal rats cause brain cell death and have significant adverse impacts on brain development and behavioral activities in adolescence. Specifically, rats that suffer painful inflammatory nociception in infancy developed hyperalgesia and abnormal stress behaviors. We show for the first time that rhEPO is an effective drug that antagonizes the inflammatory response and brain damage caused by inflammatory pain. rhEPO treatment prevents the development of hyperalgesia and corrects the stressful behavior chronically developed in these young animals. Moreover, treatment with rhEPO maintains normal brain and body weights that were reduced in animals suffering repetitive inflammatory stimuli.
Contrary to what was previously believed neonates can feel pain and, in many cases, pain is more intense than in adults due to the underdeveloped inhibitory mechanisms in the brain and spinal cord [1, 7, 8]. In clinical cases, painful stimuli and inflammatory pain occur in human infants. For example, prematurely born infants in the intensive care unit are subjected to painful procedures including, but are not limited to, heel sticks, endotracheal intubations, respiratory and gastric suctioning and catheter insertions . It is noticed that these infants may develop behavioral and emotional problems in childhood, altered pain responses, depression and other psychiatric disturbances in adulthood . The cellular, molecular and pathological mechanisms of this disorder are not well understood. Previous studies showed that full term neonates exposed to painful stimuli at birth develop acute adverse effects that last only for hours and are not chronic like those seen in premature babies [42, 43]. It is likely that the data from the present investigation are more relevant to immature infants subjected to severe and repetitive inflammatory pain.
Formalin s.c. injection is used as a model for inflammatory pain. Formalin is usually applied as a fixative in immuno-histochemical staining. It fixes tissues by reversibly cross-linking primary amino groups in proteins with nearby nitrogen atoms. When injected subcutaneously, formalin causes extensive firing of primary afferent nerves such as the C-fibers and widespread signaling activities in the brain and spinal cord  resulting in a state of repetitive inflammatory pain in animals . Formalin causes extensive neuronal activity that lead to glutamate excitotoxicity and increased neuronal death in the brain . In line with these pathophysiological events, we observed high levels of TUNEL and FJC staining in the cortex, hippocampus and the hypothalamus after formalin injection. The cortical regions are the primary somatosensory and associative sensory areas, which receive input from the primary afferents through the thalamus. The hippocampus, on the other hand, is part of the limbic system and it extensively receives nociceptive input and mediates a variety of functions including emotions and behaviors, whereas the hypothalamus, part of the hypothalamic-pituitary-adrenal axis, mediates a variety of metabolic, neuroendocrine and autonomic activities in response to physiologic and pathologic stresses. The roles of these areas in mediating somatosensory and behavioral functions and excitotoxicity explain, at least partly, the consequences of the high cell death levels detected after formalin injection. A recent study showed that formalin stimulation in neonatal rats causes long-term behavioral changes, altering exploratory behavior and anxiety levels in adulthood in a gender (female) specific manner . In that study, a relatively minor insult of single paw formalin injection was tested. It will be interesting to see if gender plays a role in the severe inflammatory pain model.
Previous work focused on the use of anesthetics and a few other non-pharmacological interventions to reduce pain and to prevent the emergence of long-term complications of early exposure to painful stimuli. However, given the controversial role and adverse effects of various anesthetics on the immature brain, for instance induction of neuronal apoptosis, these treatments have raised increasing concerns. The present investigation is the first attempt to utilize a neuroprotective agent, rhEPO, to prevent the long term complications of early painful experiences in neonates. rhEPO has been studied for its safety and efficacy in neonatal brain injury and for stimulating erythropoiesis in human infants. For example, extremely low birth weight infants who were given rhEPO at birth to stimulate erythropoiesis showed better mental capacities and improved developmental assessment at the age of 10 [47–49] although other reports failed to show neurodevelopmental benefits with rhEPO treatment . It is important to mention, however, that the neuroprotective effects of rhEPO are achieved at doses higher than those required for its erythropoietic actions. Nonetheless, human infants did not develop any of the complications associated with rhEPO seen in adult patients like hypertension, clotting, polycythemia, and tumor angiogenesis [51, 52]. Moreover, the erythropoietic actions of EPO are beneficial in preterm infants who are mostly anemic. We have also shown that rhEPO injections do not affect cytokine expression in control animals and that neonatal rhEPO treatment does not modify behaviors in adolescence. rhEPO was tested in this investigation as a preventative and protective treatment for neonates subjected to severe painful stimuli that cause adverse impacts on CNS cells and brain development. As the injury is predicted to occur and as rhEPO is a relatively safe drug, we believe it is feasible to use rhEPO to co-treat or pre-treat these neonates. Future studies may be performed to evaluate the therapeutic benefits of delayed treatment of rhEPO in older animals.
Subcutaneous formalin injections also cause non-specific tissue damage and activate other afferents that are not related to pain processing . We cannot exclude that rhEPO may show a local anti-inflammatory effect at the site of formalin injections. We can confirm, however, that, based on the results of locomotion tests, the behavioral alterations in the formalin injected rats were not due to injured paws and impaired locomotor abilities. We have also demonstrated that rhEPO helps to restore brain and body weights compared to the formalin group. This is a likely indication of complex interacting factors including cell death, cell proliferation/regeneration and feeding behavior that is linked to the limbic/hypothalamic interaction.
EPO receptor (EPOR) is abundantly expressed in brain capillaries, and EPO can cross the blood brain barrier via endocytosis . In neonatal rat brains, rhEPO reaches significant levels at 4 hours but peaks later at 10 hours after i.p. injections . This explains the protective effect of rhEPO in reducing TUNEL and FJC staining in several brain regions. It is worth to mention that, when investigated in P21 juvenile rats, we could not identify any difference in the number of dying neurons among the three groups. This observation agrees with the idea that cell death that occurred during early brain development is critical. rhEPO has anti-apoptotic functions as shown in the reduction of cleaved caspase-3 in the present study. We additionally revealed that the inflammatory pain-triggered neuronal apoptosis is caspase-dependent but not AIF-dependent, since this caspase-independent apoptotic factor was not increased by formalin injections. EPO and EPOR are expressed in both neurons and glial cells in the central nervous system . The binding of EPO to its receptor initiates a cascade that activates the ERK-1/-2, Akt and JNK-1/-2 signaling pathways . These pathways eventually activate factors that inhibit apoptosis, decrease inflammation, increase neurogenesis and angiogenesis. EPO and EPOR also induce neurogenesis after stroke , early during embryonic development  and in vitro via Jak2/Stat3 and PI3K/AKT pathway activation . We suggest that the therapeutic benefits of rhEPO treatment may be partly attributed to an increase in brain neurogenesis as well.
Several reports in the literature attribute an anti-inflammatory action to rhEPO in stroke and endotoxin-mediated inflammation [24, 60]. Both TNF-α and IL-1β are pro-inflammatory  whereas IL-6 can either be pro- or anti-inflammatory . In the present investigation, rhEPO reduced IL-1β, IL-6 and TNF-α at many time points whereas it increased TNF-α at the 4 hrs time point. It is worth to note that IL-6 significantly dropped in the formalin group while it increased in the rhEPO treated group along the 3 time points (from 4 to 72 hrs). TNF-α, on the other hand, significantly increased in the formalin group while it dropped in the rhEPO treated group in the same time span. In addition, it is well established that levels of SP and CGRP drop in the DRG in models of chronic pain  and in the spinal cord in a neonatal pain model . We demonstrate, similar to other reports , that rhEPO can reverse signs of central sensitization (restoring normal levels of SP and CGRP). CGRP expression is regulated by an upstream MAPK-responsive enhancer element. Since EPO activates the MAPK signaling cascade, the increase in CGRP expression can be attributed to EPO-induced activation of this upstream MAPK-responsive enhancer . Only mRNA levels were measured in this investigation, whether the protein levels of these factors are similarly changed remains to be confirmed.
A recent report shows that rhEPO enhances cognitive function of memory and learning . This is in line with our observations in current investigation and, collectively, it is suggested that rhEPO and EPOR may be explored as targets in a preventive therapy for premature neonates subjected to repeated painful procedures.